Mailing-List: contact dev-help@opennlp.apache.org; run by ezmlm
Precedence: bulk
Reply-To: dev@opennlp.apache.org
From: Daniel Russ <danruss00@gmail.com>
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: quoted-printable
Mime-Version: 1.0 (Mac OS X Mail 10.3 \(3273\))
Subject: Re: Early stopping NameFinderME
Date: Fri, 25 Aug 2017 19:07:03 -0400
References: <CAJX40TFjHk5dCpngF2x+O8tHUCwRifOJrLdbyMvcz3M8mnOkFg@mail.gmail.com>
 <CA+V=Wqipe3SVdu5yt9YjuuD_ug3CWtPiWD0P+4PrPXr9knDE1g@mail.gmail.com>
To: dev@opennlp.apache.org
In-Reply-To: <CA+V=Wqipe3SVdu5yt9YjuuD_ug3CWtPiWD0P+4PrPXr9knDE1g@mail.gmail.com>
Message-Id: <AB2DF928-3853-4323-8453-E3C6DAC321FC@gmail.com>
archived-at: Fri, 25 Aug 2017 23:07:12 -0000

J=C3=B6rn,

   Currently, GISTrainer has a private static final variable =
LLThreshold, which controls if the change in the log likelihood between =
two iterations is too small.  We could make this parameter. I am =
concerned about using the accuracy to train the model.  If we use =
accuracy, the weight space may be flat. =20

   Saurabh, you use the term =E2=80=9Cearly stopping=E2=80=9D.  In deep =
learning, early stopping is used to prevent overtraining and improve =
generalization to unseen data.  I am not sure early stopping serves the =
same purpose with GIS training.  Does anyone know if early stopping =
improves generalization for a maxent problem?

Daniel

> On Aug 24, 2017, at 4:48 AM, Joern Kottmann <kottmann@gmail.com> =
wrote:
>=20
> You are the first one who ever asked this question. I think we have =
this as
> an option already on the gis trainer but it is not exposed all the way
> through.
>=20
> Please open a jira and I can look at it next week.
>=20
> J=C3=B6rn
>=20
> On Aug 21, 2017 5:11 PM, "Saurabh Jain" <saurabh4768jain@gmail.com> =
wrote:
>=20
>> Hi All
>>=20
>> How can we use early stopping while training/crossvalidating custom =
data
>> with NameFinder ? What I want if change in likelihood value or =
accuracy of
>> model is less than 0.05 between two steps (differ by 5 i.e compare =
x+5 step
>> output with x step) then training should stop. I could not find =
anything
>> regarding this in documentation. Can some one please help ?
>>=20
>> --
>> *Thanks & Regards*
>>=20
>>=20
>> *Saurabh Jain *
>> *AI Developer*
>>=20
>> *Active Intelligence  *
>>=20
>> *"*
>> *To do a thing yesterday was the best time . Second best time is =
today .=E2=80=9D *
>>=20